Ivan Dokuchaev

Working with a leading elevator industry client, providing expertise in data science to support designing and implementing machine learning solutions in predictive maintenance. - Designed and developed an elevator fault classification system, achieving over 20% performance improvement compared to the previous implementation. - Processed and analyzed highly unstructured data using Pandas and Pyspark, applying advanced data cleaning techniques and feature engineering to enhance model performance.

- Developed an LSTM neural network-based machine learning model to estimate fracture dimensions following hydraulic fracturing operations. - Conducting feature engineering on extensive wellbore sensor data in HDF5 binary format, utilizing H5py and Pandas for processing and analysis. - Provided valuable insights for stakeholders by creating and showcasing time series data visualizations utilizing Matplotlib.

Master project: Title: “Android malware families and their impact on ML-based malware classifiers”

Bachelor project: Title: “Deep Learning-based Person Biometric Identification”